Plant defense against pathogens is induced by R-genes, a large multigene family of signal transducing proteins. Approximately a dozen R-genes have been shown to be segregating for both resistance and susceptibility alleles, and several of these are known to harbor ancient polymorphisms. R-genes can also show evidence of rapid, adaptive evolution, presumably as a consequence of coevolutionary interactions with pathogens. That both of these evolutionary processes -balancing selection and adaptive evolution -occur in the same gene is remarkable, and requires explanation. The general aims of this proposal are to elucidate the mechanisms of natural selection acting on segregating R-gene alleles and to determine the physiological basis for their fitness differences. We believe that both these pursuits are necessary to fully understand the evolutionary genetics of disease resistance variation. Sequence variation at three complex loci containing a variable number of R-gene repeats will also be investigated. We hypothesize that complex loci evolve in response to strong balancing selection and large costs of specific resistance alleles. The data will allow us to evaluate alternative hypotheses, and potentially to unite simple and complex R-gene loci under a single model of selection. Our specific aims in this grant period will be to (1) Test the hypothesis that a large cost of resistance is characteristic of long-lived polymorphisms between resistance and null susceptibility alleles; (2) Test the prediction that this cost is lower for alleles that participate in a stable polymorphism between functional alternatives; (3) Analyze the cost of Rpml resistance by characterizing R and S lines in terms of gene expression, biochemical intermediates and biochemical end products; (4) Test the role of particular defense responses in generating the cost of Rpml resistance; (5) Characterize and contrast nucleotide polymorphism of single-copy R-gene loci in a reference set of 96 A. thaliana accessions, and of tandem arrays in a subset of these samples, to identify the spectrum of evolved responses by R-genes to disease; and (6) Develop population genetic models of the major modes of R-gene polymorphism.